A Systematic Review of Selective and Non-Selective Beta Blockers for Prevention of Vascular Events in Patients with Acute Coronary Syndrome Or Heart Failure

Home , Bucindolol

REVIEW A systematic review of selective and non-selective beta blockers for prevention of vascular events in patients with acute coronary syndrome or heart failure

O.R. de Peuter1*, F. Lussana1,2, R.J.G. Peters3, H.R. Büller1, P.W. Kamphuisen1

1Department of Vascular Medicine and 3Cardiology, Academic Medical Centre, University of Amsterdam, Amsterdam, the Netherlands, 2Unit of Hematology and Thrombosis, Department of Medicine, Surgery and Dentistry, Ospedale San Paolo, University of Milan, Milan, Italy, *corresponding author: tel.: +31 (0)20-566 75 16, fax: +31 (0)20-566 93 43, e-mail: [email protected]

A b s t r a c t

Background: To assess the influence of b2-receptor mortality and vascular events in patients with ACS or, to suppression on top of selective b1-receptor blockade on the a lesser extent, HF. However, only a few studies directly occurrence of vascular events and on all-cause mortality compared b blockers, and indirect comparisons were in patients with acute coronary syndrome (ACS) or heart subject to heterogeneity, which weakens firm conclusions. failure (HF). Methods: Systematic review of studies published since 1980. Randomised controlled trials directly comparing Keywords b1 blockers with b1+2 blockers, or comparing the two b blockers with placebo, were included. Studies had a Epidemiology, heart failure, myocardial infarction, minimum treatment period of three months and total pharmacology, prevention mortality or vascular events as their primary or secondary outcome. Results: Of the included studies, five directly compared Introduction b blockers (3733 patients) and 28 compared b blockers with placebo (30,889 patients). These latter studies were Beta-adrenergic receptor blocking agents (b blockers) are heterogeneous in study population, dose and type of generally recommended for the treatment of patients with b blockers. In ACS, the only study directly comparing acute coronary syndrome (ACS) or heart failure (HF), different b blockers was underpowered to detect a difference because of their proven positive effects on life expectancy, on mortality, while in HF b1+2 blockers significantly risk of sudden cardiac death and left ventricular ejection decreased mortality compared with b1 blockers (RR 0.86, fraction.1-4 In patients with HF, b blockers inhibit the 95% confidence interval 0.78 to 0.94). In ACS, b1 blockers adverse effects of an increased sympathetic activity, which in placebo-controlled trials non-significantly reduced total has been associated with increased mortality.2,5 mortality (RR 0.82, 0.67 to 1.01) or vascular events (RR Beta blockers can be classified as b blockers with a much 0.68, 0.42 to 1.11), while b1+2 blockers were associated with higher affinity for b1- than for b2-adrenergic receptors (b1 a significant decrease in total mortality (RR 0.73, 0.64 to blockers), and b blockers with both b1- and b2-adrenergic 0.82), and vascular events (RR 0.71, 0.59 to 0.84). In HF, receptor blocking properties (b1+2 blockers).6 Previous b1 and b1+2 blockers reduced total mortality, while only meta-analyses have suggested a better effect of b1+2 blockers b1+2 blockers decreased vascular events (RR 0.80, 0.64 on total mortality and cardiovascular morbidity in patients to 1.00). with ACS7 and HF,8,9 although these parameters were not Conclusions: Additional b2-receptor blockade may be more the primary outcomes in these trials. Furthermore, in the effective than b1-receptor blockade alone in preventing total large COMET trial, carvedilol (a b1+2 blocker) significantly

october 2009, vol. 67, no 9 284 reduced cardiovascular mortality compared with the b1 publication with the data that best corresponded to our blocker metoprolol in patients with HF.10 Interestingly, the objectives and supplemented it, if necessary, with data reduction in cardiovascular mortality was largely driven by a from the other publications. To assess the agreement difference in vascular events.11 The underlying mechanism between reviewers for study selection, we used the kappa of this effect is unclear. One hypothesis is that sympathetic (κ) statistic, which measures agreement beyond chance.17 activity may influence vascular events by increasing the We included randomised controlled or active controlled prothrombotic activity.12,13 This sympathetic activity may be trials that directly compared b1 and b1+2 blockers. Since reduced by a specific presynaptic b2-adrenergic inhibitory there were only a few trials comparing these compounds effect of b1+2 blockers,14-16 resulting in less activation of directly, we also assessed randomised placebo-controlled platelets and clotting factors.13 Since patients with HF trials. Patients with systolic heart failure were included have an increased sympathetic activation, b blockers with regardless of the underlying cause of heart failure or b2-adrenergic inhibitory effects could reduce the associated cardiac rhythm. Only studies with prespecified outcomes prothrombotic activity and, consequently, the number of of mortality or vascular events were included. To assess the vascular events. long-term effects of b-blocker treatment only studies with We therefore performed a systematic review of all at least three months of treatment were considered. Studies randomised studies assessing b blockers in patients with assessing b-blockers with intrinsic sympathicomimetic, ACS and HF to test the hypothesis that suppression of the class-III antiarrhythmic or partial agonist activity, were b2-adrenergic receptor in addition to the b1-adrenergic excluded. The oldest trial with b1 blockers was published receptor is more effective in reducing vascular events than in 1981.18 To facilitate a balanced comparison between the a more selective suppression of the b1 receptor. different b blockers, only studies published since that time were included.

Methods Data extraction and quality assessment Using a data extraction form, two authors independently Study selection extracted the following baseline characteristics for all To test our hypothesis we divided b blockers into b included studies: first author, year of publication, source blockers that antagonise the b1 receptor more selectively (b1 of publication, country of origin, study design, inclusion blockers), and b blockers with both b1- and b2-adrenergic and exclusion criteria, type of b blocker used and dosage, receptor blocking capacities (b1+2 blockers). The effects of concomitant medication, duration of follow-up, cause of b1 and b1+2 blockers for secondary prevention in patients heart failure, NYHA classification and left ventricular with ACS or HF were analysed. The primary outcomes ejection fraction, number, mean age and gender of the evaluated were 1) all-cause mortality and 2) vascular events, study patients. The following outcomes were retrieved: defined as fatal and non-fatal strokes, fatal and non-fatal all-cause mortality and number of strokes, myocardial myocardial infarctions and fatal pulmonary embolisms and (re)infarctions, or venous thromboembolic events (all other venous thromboembolic events. fatal and non-fatal events). Since the numbers of reported We conducted a comprehensive literature search of Medline, events were relatively small, we analysed these events EMBASE and the Cochrane Central Register of Controlled as a composite endpoint. For each study the number Trials library from 1981 to June 2009. In Medline text and of patient-years was calculated by multiplying the total Cochrane library keywords were “randomised controlled number of patients with the mean follow-up period. trial”, “acute coronary syndrome” (in Cochrane “myocardial Study quality was evaluated as described by Jadad et ischemia”) or “congestive heart failure” and “adrenergic al.19 Studies using adequate treatment allocation beta-antagonists”, using Medical Subject Heading Terms. sequence, proper concealment, blinding of both patient In EMBASE text keywords were “randomised controlled and investigator, and completeness of follow-up, were trial”, and “heart muscle ischemia” or “heart failure”, and considered to reflect higher methodological quality. Since “beta adrenergic receptor blocking agent”. The results of not all studies provided information on these quality the searches were limited to studies of humans and were criteria, we also assessed the effect of including only trials not restricted to English language. In addition a review of with adequate concealment and loss to follow-up <20%. references from primary or review articles was performed to identify any additional relevant studies. Statistical analysis The list of articles was reviewed by two authors, who There were only a few studies that directly compared b1 independently evaluated all articles for possible inclusion. and b1+2 blockers. Therefore, despite potential biases of Disagreement was resolved by consensus and if necessary comparisons between different studies, the effects of b1 by the opinion of a third reviewer. When multiple papers and b1+2 blockers were indirectly analysed by pooling the for a single study had been published, we used the results of placebo-controlled studies. Relative risks (RR)

De Peuter, et al. Beta blockers and vascular events.

october 2009, vol. 67, no 9 285 and 95% confidence intervals (95% CI) were calculated Figure 1. Results of article search and selection using the DerSimonian and Laird random-effects model20 with the Review Manager developed by the Potentially relevant Cochrane Collaboration, version 4.2.10 for Windows. A studies identified and screened for random-effects model was chosen since results with this retrieval (n=4258) model are more conservative. Statistical heterogeneity Studies excluded 2 2 after title and between studies was evaluated using the χ and I test, abstract screening for each of the outcomes separately, with a p value <0.05 using the predefined considered as heterogeneous. inclusion criteria (n=4173) Sensitivity analyses were performed to evaluate the Studies retrieved robustness of the results. First, we considered the effect for detailed evalu of only including high-quality studies with a double-blind ation (n=85) design and with a number of loss to follow-up less than Studies excluded after full article screening 20%. In addition, the effect of excluding studies one at Studies found by inclusion criteria a time to identify those that may have a disproportionate manual reviewed of (n=48): reference list (n=2) • Duplicate data (n=25) influence on the summary treatment effect was evaluated. • Outcomes (n=12) Studies included in Publication bias was assessed using a funnel plot of effect • Follow-up less than 3 the review (n=33) 21 months (n=1) size versus standard error. In a final analysis the results Duplicates with • Diastolic heart additional informa- were adjusted for number of patient-years to assess the failure (n=1) tion (n=6) influence of duration of study follow-up. • Stage II of a trial with other medications (n=1) • Patients with Results suspected myocardial infarction or chronic coronary artery Literature search disease (n=6) • Short treatment Figure 1 summarises the process of study selection. A duration (n=1) total of 4258 relevant literature citations were identified, • Summary report (n=1) of which 4173 were excluded after scanning titles and abstracts, leaving 85 studies for detailed assessment. Two additional studies were identified through a manual review of study bibliographies.22,23 Of these 87 retrieved patients with HF.10,37,41,44 Twenty-eight studies compared articles, 48 were excluded for the following reasons: 12 a b1 or b1+2 blocker with a control group, of which 11 because total mortality or vascular events were not a studies enrolled patients with ACS18,22,23,27,28,35,39,42,45,50,51 prespecified outcome of the study; one study because and 17 patients with HF.24-26,29-34,36,38,43,46-49,52 The number patients were followed up for less than three months; 25 of patients among the studies ranged from 50 to 3991. because of duplicate data, substudies or commentaries; six In one study with three treatment arms in different studies were excluded because of inclusion of patients with dosages;29 we only included the group given the target suspected myocardial infarction or chronic coronary artery dose used in most other studies. In another study the diseases, without documented ACS; one study because it results from 326 of 764 subjects were excluded, since included patients with diastolic heart failure; one because these patients had no acute myocardial infarction or it was the summary report of four randomised trials with HF.23 carvedilol; another because the treatment with b blockers Various b blockers were studied: the b1 blockers was the second part of a trial with other medications; included metoprolol, nebivolol, bisoprolol, atenolol and and one because treatment duration was only seven days. betaxol and the b1+2 blockers included carvedilol, Hence, of the 39 remaining studies 33 were included in bucindolol, propranolol and timolol. In trials that the present systematic review,10,18,22-52 with a total of 34,360 included information on concomitant treatment, on patients (table 1). Six additional studies reported additional average 91% of the patients received angiotensin- information.11,53-57 The interobserver agreement for study converting enzyme (ACE) inhibitors, 91% diuretics, selection was excellent (κ =0.98). and 74% digitalis. No information was available on statin use. Twenty-seven studies were reported as Study characteristics and quality double blind,10,18,22,23,26-39,42-46,48-50,56 13 had appropriate Table 1 shows the study characteristics of the 33 trials random allocation of treatment10,26,30-32,35,38-40,43,49,51,52 and included. Five studies directly compared b1 with b1+2 in ten studies information on concealment allocation blockers, one assessing patients with ACS40 and four was adequate.10,26,30-32,35,38,43,49,51 A description of patient

De Peuter, et al. Beta blockers and vascular events.

october 2009, vol. 67, no 9 286 Table 1. Study and participant summary characteristics

Year Total Mean Male Participants Treatment Target dose Main concomi- Average Patient- (n) age (%) tant medication follow- years (years) up in months Patients with ACS; direct comparator b blocker CAMIS40 2005 232 61 78 Within 24 hours Carvedilol/ 25 mg bid/ Aspirin, statin, 18 348 after AMI Atenolol 50 mg bid vasodilator Patients with HF; direct comparator b blocker BETACAR39 2006 255 57 86 NYHA II-III, LVEF Carvedilol/ 25 mg bid/ Vasodilator, 8 170 <35% Betaxolol 20 mg od diuretic, digitalis, nitrates COMET10 2003 3029 62 80 NYHA II-IV, LVEF Carvedilol/ 25 mg bid/ Vasodilator, 58 14640 <35% Metoprolol 50 mg bid diuretic, digitalis Kukin41 1999 67 58 69 NYHA II-IV, LVEF Carvedilol/ 25 mg bid‡ / Vasodilator, 6 34 <35% Metoprolol 25 mg bid‡ diuretic, digitalis Metra44 2000 150 57 91 NYHA II-IV, LVEF Carvedilol/ 25 mg bid†/ Vasodilator, 14 175 <35% Metoprolol 50 mg bid† diuretic, digitalis Patients with ACS; b1 blockers Goteborg8 1983 1395 66% 76 AMI Metoprolol 100 mg bid None reported 3 349 <65 years Lopressor42 1987 2395 58 83 5-15 days after AMI Metoprolol 100 mg bid None reported 12 2395 Manger Cats22 1983 553 100% ? <1 year after AMI Metoprolol 100 mg bid None reported 12 553 <70 years Olsson45 1985 301 60 81 Within 2 days after Metoprolol 100 mg bid Digitalis, 36 903 AMI diuretics Salathia23 1985 474 6% 71 AMI Metoprolol 100 mg bid None reported 12 474 <65 years Patients with ACS; b1+2 blockers BEAT51 2002 343 69 83 Within 7 days after Bucindolol 50 mg bid† Vasodilator, 7.5 214 AMI, LVEF <35% diuretic, digitalis BHAT28 1982 3837 55 84 5-21 days after AMI Propranolol 60-80 mg None reported 25 7994 3/day Basu27 1997 151 60 81 AMI Carvedilol 25 mg bid Aspirin, heparin, 6 76 thrombolysis, nitrates CAPRICORN35 2001 1959 63 74 3-21 days after AMI, Carvedilol 25 mg bid Vasodilator, 16 2612 LVEF <40% diuretics, aspirin Hansteen55 1982 560 58 85 4 days after AMI Propranolol 40 mg 4/ None reported 12 560 day Pedersen57 1983 1884 61% 62 6-27 days after AMI Timolol 10 mg bid Diuretics, 17 2669 <65 years digitalis Patients with HF; b1 blockers Anderson24 1985 50 51 66 LVEF <40%, Metoprolol 50 mg bid Vasodilator, 19 79 idiopathic diuretic, digitalis, anticoagulant CIBIS-I31 1994 641 60 83 NYHA III-IV, LVEF Bisoprolol 5 mg od Vasodilator, 23 1229 <40% diuretic, digitalis (in 56% of patients) CIBIS-II30 1999 2647 61 81 NYHA III-IV, LVEF Bisoprolol 10 mg od Vasodilator, 16 3529 <35% diuretic ENECA36 2005 260 72 73 NYHA II-IV, LVEF Nebivolol 10 mg od Vasodilator, 8 173 <35% diuretic, digitalis SENIORS38 2005 2128 76 63 LVEF <35% Nebivolol 10 mg od Vasodilator, 21 3724 diuretic MERIT-HF43 1999 3991 64 78 NYHA II-IV, LVEF Metoprolol* 200 mg od Vasodilator, 12 3991 <40% diuretic Waagstein52 1993 383 49 ? LVEF <40%, idio Metoprolol 100-150 mg Vasodilator, 12 383 pathic dilated 2-3/day diuretic, digitalis cardiomyopathy This table continues on the next page.

De Peuter, et al. Beta blockers and vascular events.

october 2009, vol. 67, no 9 287 Table 1. Continued

Year Total Mean Male Participants Treatment Target dose Main concomi- Average Patient- (n) age (%) tant medication follow- years (years) up in months Patients with HF; b1+2 blockers Aranow25 1997 158 81 29 NYHA II-III, LVEF Propranolol 30 mg 3/ Vasodilator, 32 421 >40%, prior ACS day diuretic, digitalis in AF Austr/NZ HF26 1997 415 67 80 NYHA II-III, LVEF Carvedilol 25 mg bid Vasodilator, 19 657 <45%, ischaemic diuretic, digitalis cause BEST49 2001 2708 60 78 NYHA III-IV, LVEF Bucindolol 50-100 mg Vasodilator, 24 5416 <35% bid diuretic, digitalis CHRISTMAS32 2003 387 63 90 NYHA I-III, LVEF Carvedilol 25 mg bid Vasodilator, 6 194 <40%, ischaemic diuretic cause COPERNICUS47 2001 2289 63 80 LVEF <25% Carvedilol 25 mg bid Vasodilator, 10 1908 diuretic, digitalis Cohn33 1997 105 60 69 <150 meter on Carvedilol 25 mg bid Vasodilator, 6 53 walking test, LVEF diuretic, <35% digitalis, nitrates Colucci34 1996 366 54 85 425-550 meter Carvedilol 25 mg bid‡ Vasodilator, 7 214 on walking test, diuretic, LVEF<35% digitalis, nitrates MOCHA29 1996 173 60 77 150-425 meter on Carvedilol 25 mg bid Vasodilator, 6 87 walking test, LVEF diuretic <35% Palazzuoli48 2005 58 71 66 NYHA III-IV, LVEF Carvedilol 25 mg bid Vasodilator, 12 58 <40% diuretic, digitalis, anticoagulant PRECISE46 1996 278 60 73 150-450 meter on Carvedilol 25 mg bid‡ Vasodilator, 6 139 walking test, LVEF diuretics, <35% digitalis

ACS = acute coronary syndrome; HF = heart failure; AMI = acute myocardial infarction; LVEF = left ventricular ejection fraction; NYHA = New York Heart Association; B = blinding of outcome measure; R = randomisation; bid = twice daily, od = once daily; P = one of our endpoints was a primary outcome. W = description of withdrawals. ‡If bodyweight >85 kg target dose was doubled. †If bodyweight >75 kg target dose was doubled. * Metoprolol CR/XL. withdrawal was provided in all studies except two.22,37 Indirect comparison of b1 and b1+2 blockers in ACS Based on Jadad’s scale,19 ten out of 33 (30.3%) studies In the five studies on b1 blockers,18,22,23,42,45 treatment were rated as high quality.10,26,30-32,35,38,39,43,49 resulted in a non-significant reduction of all-cause mortality compared with placebo (RR 0.82, 95% CI 0.67 to Direct comparison of b1 and b1+2 blockers in patients with 1.01) ( figure 3). Information on vascular events was available ACS and HF in three of the five studies.18,42,45 Fifty-six of the 2047 (3%) For ACS, the only study with direct comparison of different patients with b1 blockers and 81 of 2044 (4%) patients in b blockers (n=232) showed no difference on all-cause the control group had vascular complications, without a mortality (RR 0.39, 95% CI 0.08 to 1.95).40 No data were statistical significant difference between the two groups available on vascular events. (RR 0.68, 95% CI 0.42 to 1.11) (figure 3). There was no clear In four studies that directly compared the effects of b1 heterogeneity across studies for both outcomes (I2=31.4 %, and b1+2 blockers on total mortality in patients with p=0.20 and I2=0 %, p=0.61, respectively). HF,10,37,41,44 b1+2 blockers significantly decreased total Compared with placebo, b1+2 blockers reduced total mortality compared with b1 blockers (RR 0.86, 95% CI 0.78 mortality in patients with ACS, (RR 0.73, 95% CI 0.64 to to 0.94) (figure 2). It should be noted that the COMET trial10 0.82).27,28,35,39,50,51 In addition, b1+2 blockers also lowered contributed to more than 96% of these results. Only the the risk of vascular events by 29% (RR 0.71, 95% CI 0.59 COMET trial reported vascular mortality and morbidity,11 to 0.84), occurring in 395 of 4361 (9%) patients with b showing a significantly better effect of the b1+2 blocker blockers and in 545 of 4373 (12%) patients with placebo carvedilol in reducing fatal and non-fatal myocardial (figure 3). This effect was consistent in all studies except infarction and death from stroke (HR 0.70, 95% CI 0.50 to one.39 There was no significant statistical heterogeneity 0.99, and HR 0.33, 95% CI 0.18 to 0.62, respectively). among the studies (I2=0%, p=0.79).

De Peuter, et al. Beta blockers and vascular events.

october 2009, vol. 67, no 9 288 Figure 2. Direct comparison of b1 and b1+2 blocker treatment on total mortality

Patients with acute coronary syndrome (total mortality) Study b1+2 b1 blockers Relative risk (random) Weight (%) Relative risk blockers (n/N) (95% CI) (n/N) CAMIS 200541 21/118 5/114 100.00 0.39 (0.08, 1.95) Total 21/118 5/114 100.00 0.39 (0.08, 1.95) 0.1 0.2 0.5 1 2 5 10 Test for heterogeneity: not applicable Test for overall effect: Z = 1.15 (p=0.25)

Patients with heart failure (total mortality) BETACAR 200640 7/131 3/124 0.48 2.21 (0.58, 8.35) COMET 200310 512/1511 600/1518 96.58 0.86 (0.78, 0.94) Kukin 199942 2/37 1/30 0.15 1.62 (0.15, 17.03) Metra 200045 17/75 21/75 100.00 0.86 (0.78, 0.94) Total 538/1754 625/1747 100.00 0.86 (0.78, 0.94) Test for heterogeneity: c2 = 2.27, df = 3 (p=0.52), 12 = 0% Test for overall effect: Z = 3.18 (p=0.001) 0.1 0.2 0.5 1 2 5 10 Favours Favours b1+2 blockers b1 blockers

95% CI = 95% confidence interval; n = number of events; N = number of included patients.

Indirect comparison of b1 and b1+2 blockers in HF the effect of different b blockers. Assessing various factors In seven studies that assessed all-cause mortality in patients in Jadad’s scale did not affect the significant associations with heart failure,24,30,31,36,38,43,52 558 of 5057 patients (11%) who in the results. Moreover, including only higher quality received b1 blockers died compared with 736 of 5043 patients studies, results were virtually the same (table 2). When (15%) in the control group, resulting in a reduction in mortality we analysed the long-term effects of b-blocker treatment (RR 0.76, 95% CI 0.68 to 0.84) (figure 4). Data on vascular among studies that had a follow-up of at least 12 months, events were available in three placebo-controlled trials,30,31,52 the overall results on total mortality and on fatal and involving 3671 patients. In these studies, b1 blockers did not non-fatal vascular events did not change. Due to the low protect patients against vascular events, as compared with number of reported events, separate analysis for myocardial placebo (RR 1.33, 95% CI 0.86 to 2.04) (figure 4). There was no infarction and stroke was not possible. significant heterogeneity for both outcomes (I2=0 %, p=0.38 The funnel plots for the studies with patients with ACS were and I2=19.3 %, p=0.28, respectively) (figure 4). symmetrical, indicating no publication bias. There was some For b1+2 blockers, ten trials assessed all-cause asymmetry in the funnel plots for b blockers in patients with mortality,25,26,29,32-34,46-49 625 of 3546 (18%) HF patients who HF, indicating a possible publication bias. Because of its received b1+2 blockers died, compared with 762 of 3391 low power and relatively small number of included trials (22%) patients in the control group, with a statistically (maximum of ten included studies per group), we did not significant reduction in mortality (RR of 0.75, 95% CI 0.61 perform an Egger’s regression analysis. Correcting the to 0.92) (figure 4). For this latter outcome there was some results for number of patient-years provided similar pooled heterogeneity across the studies (I2=41%, p=0.04). relative risks for all figures (data not shown). Six trials reported vascular events in HF.25,26,29,46,47,49 Compared with placebo, b1+2 blockers were associated with a 20% decrease in vascular events (RR 0.80, 95% CI Discussion 0.64 to 1.00). All trials reported fatal or non-fatal events of worsening This systematic review confirms the beneficial effects ofb heart failure; b1 and b1+2 blockers equally decreased these blockers on total mortality, as reported previously.8,9,58 We events (RR 0.77, 95% CI 0.61 to 0.97, and RR 0.82, 95% CI specifically assessed the beneficial effect of b2-adrenergic 0.70 to 0.95 respectively, data not shown). receptor blockade on vascular events in patients with ACS or HF. Beta blockers with b2-adrenergic inhibitory effects Sensitivity analysis could reduce sympathetic activation and the associated The results of our primary analyses were unaffected by prothrombotic activity, and, consequently, the number of removing individual studies one by one or by analysing vascular events. Indeed, our results suggest a somewhat

De Peuter, et al. Beta blockers and vascular events.

october 2009, vol. 67, no 9 289 Figure 3. Effects of b1 and b1+2 blocker treatment on total mortality and vascular events in patients with acute coronary syndrome

Total mortality Study Treatment Control RR (random) Weight (%) RR (95% CI) n/N n/N b1 blockers Goteborg19 40/698 62/697 7.05 0.64 (0.44, 0.95) Lopressor trial43 65/1195 62/1200 9.03 1.05 (0.75, 1.48) Manger Cats23 9/273 16/280 1.62 0.58 (0.26, 1.28) Olsson46 25/154 31/147 4.57 0.77 (0.48, 1.24) Salathia24 38/250 38/224 6.10 0.90 (0.59, 1.35) Subtotal 177/2570 209/2548 28.37 0.82 (0.67, 1.01) Test for heterogeneity: χ2 = 4.59, df = 4 (p = 0.33), I2 = 12.9% Test for overall effect: Z = 1.85 (p = 0.06) b1+2 blockers BEAT52 27/170 30/173 4.59 0.92 (0.57, 1.47) BHAT29 138/1916 188/1921 23.39 0.74 (0.60, 0.91) Basu28 2/77 3/74 0.33 0.64 (0.11, 3.73) CAPRICORN36 116/975 151/984 20.42 0.78 (0.62, 0.97) Hansteen56 25/278 37/282 4.50 0.69 (0.42, 1.11) Pedersen58 98/945 152/939 18.39 0.64 (0.51, 0.81) Subtotal 406/4361 561/4373 71.63 0.73 (0.64, 0.82) Test for heterogeneity: χ2 = 2.40, df = 5 (p = 0.79), I2 = 0% Test for overall effect: Z = 5.19 (p = 0.00001)

Total 583/6931 770/6921 100.00 0.75 (0.68, 0.84) 0.1 0.2 0.5 1 2 5 10

Vascular events b1 blockers Goteborg19 9/698 15/697 3.69 0.60 (0.26, 1.36) Lopressor trial43 25/1195 24/1200 7.29 1.05 (0.60, 1.82) Olsson46 22/154 42/147 9.69 0.50 (0.31, 0.79) Subtotal 56/2047 81/2044 20.67 0.68 (0.42, 1.11) Test for heterogeneity: χ2 = 4.09, df = 2 (p = 0.13), I2 = 51.2% Test for overall effect: Z = 1.56 (p = 0.12) b1+2 blockers BEAT52 5/170 17/173 2.68 0.30 (0.11, 0.79) BHAT29 197/1916 254/1921 28.17 0.78 (0.65, 0.93) Basu28 7/77 17/74 3.68 0.40 (0.17, 0.90) CAPRICORN36 34/975 57/984 11.41 0.60 (0.40, 0.91) Hansteen56 27/278 31/282 8.93 0.88 (0.54, 1.44) Pedersen58 125/945 169/939 24.45 0.73 (0.59, 0.91) Subtotal 395/4361 545/4373 79.33 0.71 (0.59, 0.84) Test for heterogeneity: χ2 = 7.29, df = 5 (p = 0.20), I2 = 31.4% Test for overall effect: Z = 3.89 (p < 0.0001) Total 451/6408 626/6417 100.00 0.69 (0.59, 0.82)

0.1 0.2 0.5 1 2 5 10 Favours Favours treatment control RR = relative risk; 95% CI = 95% confidence interval; n = number of events; N = number of included patients.

De Peuter, et al. Beta blockers and vascular events.

october 2009, vol. 67, no 9 290 Figure 4. Effects of b1 and b1+2 blocker treatment on total mortality and vascular events in patients with heart failure Total mortality Study Treatment Control RR (random) Weight (%) RR 95% CI) n/N n/N b1 blockers Anderson25 5/25 6/25 1.10 0.83 (0.29, 2.38) CIBIS-I32 53/320 67/321 7.64 0.79 (0.57, 1.10) CIBIS-II31 156/1327 228/1320 13.17 0.68 (0.56, 0.82) ENECA37 7/134 7/126 1.16 0.94 (0.34, 2.61) MERIT-HF44 145/1990 217/2001 12.57 0.67 (0.55, 0.82) SENIORS39 169/1067 192/1061 13.19 0.88 (0.72, 1.06) Waagstein53 23/194 19/189 3.27 1.18 (0.66, 2.09) Subtotaal 558/5057 736/5043 52.11 0.76 (0.68, 0.87) Test for heterogeneity: χ2 = 7.44, df = 6 (p = 0.28), I2 = 19.3% Test for overall effect: Z = 4.24 (p = <0.0001) b1+2 blockers Aronow26 44/79 60/79ww 11.07 0.73 (0.58, 0.93) Austr / NZ HF27 20/207 26/208 3.51 0.77 (0.45, 1.34) BEST50 411/1354 448/1354 17.37 0.92 (0.82, 1.02) CHRISTMAS33 8/193 6/194 1.12 1.34 (0.47, 3.79) COPERNICUS48 130/1156 190/1133 12.24 0.67 (0.54, 0.83) Cohn34 2/70 2/35 0.34 0.50 (0.07, 3.40) Colucci35 2/232 5/134 0.47 0.23 (0.05, 1.17) MOCHA30 1/89 13/84 0.31 0.07 (0.01, 0.54) PRECISE47 6/133 11/145 1.28 0.59 (0.23, 1.56) Palazzuoli47 1/33 1/25 0.17 0.76 (0.05, 11.53) Subtotal 625/3456 762/3391 47.89 0.75 (0.61, 0.92) Test for heterogeneity: χ2 = 18.05, df = 9 (p = 0.03), I2 = 50.1% Test for overall effect: Z = 2.78 (p = 0.005)

Total 1183/8603 1498/8434 100.00 0.77 (069, 0.86) Test for heterogeneity: χ2 = 1.95, df = 2 (p = 0.38), I2 = 0% Test for overall effect: Z = 1.29 (p = 0.20) 0.01 0.1 1 10 100 Vascular events b1 blockers CIBIS-I32 10/320 11/321 7.93 0.91 (0.39, 2.12) CIBIS-II31 38/1327 24/1320 16.55 1.57 (0.95, 2.61) Waagstein53 0/194 1/189 0.67 0.32 (0.01, 7.92) Subtotal 48/1841 36/1830 23.15 1.33 (0.86, 2.04)

Test for heterogeneity: χ2 = 5.45, df = 5 (p = 0.36), I2 = 8.2% Test for overall effect: Z = 1.98 (p = 005) Aranow26 3/79 5/79 3.27 0.60 (0.15, 2.43) Austr / NZ HF27 25/207 34/208 17.64 0.74 (0.46, 1.19) BEST50 79/1354 78/1354 27.15 1.01 (0.75, 1.37) COPERNICUS48 52/1156 76/1133 24.68 0.67 (0.48, 0.95) MOCHA30 1/89 3/84 1.33 0.31 (0.03, 2.97) PRECISE47 0/133 3/145 0.78 016 (0.01, 299) Subtotal 160/3018 199/3003 74.85 0.80 (0.64, 1.00)

Total 208/4859 235/4833 100.00 0.87 (0.67, 1.14) 0.01 0.1 1 10 100 Favours Favours treatment control RR = relative risk; 95% CI = 95% confidence interval; n = number of events; N = number of included patients.

De Peuter, et al. Beta blockers and vascular events.

october 2009, vol. 67, no 9 291 Table 2. Sensitivity analysis by methodological criteria, defined as double blind and loss to follow-up less than 20%

Group of patients and outcomes b1 blockers b1+2 blockers RR (95% CI) RR (95% CI) Acute coronary syndrome: total mortality 0.84 (0.67-1.05) 0.72 (0.63-0.81) Acute coronary syndrome: vascular events 0.68 (0.42-1.11) 0.74 (0.66-0.84) Heart failure: total mortality 0.75 (0.66-0.85) 0.74 (0.56-0.96) Heart failure: vascular events 1.34 (0.82-2.18) 0.79 (0.61-1.03) RR = relative risk; CI = confidence interval.

better, and at least a more consistent, reduction of vascular and nebivolol) in both groups. These third-generation events and total mortality of additional b2-receptor blockade b blockers have additional effects such as a-receptor compared with b1-receptor blockade alone in patients with blocking properties, antioxidative effects and NO-releasing ACS. In patients with HF, b1 and b1+2 blockers both capacities and have a more favourable metabolic profile.61,62 reduced total mortality, but only b1+2 blockers had an These additional effects on our outcome parameters cannot effect on vascular events. be totally ruled out. However, the third-generation b1 Our analysis was hampered by the few trials that directly blocker nebivolol showed no clear effect on mortality and compared b1 and b1+2 blockers, including a limited vascular events. Furthermore, the beneficial effect ofb 1+2 number of patients. For ACS one trial directly compared b blockers on vascular events could have been influenced blockers, but was underpowered to detect a difference on by a1-receptor blocking properties of the b1+2 blockers mortality. For HF, four trials directly compared b blockers, carvedilol and bucindolol. However, propranolol and but these results were dominated by the large COMET timolol,28,55,57 not affecting the a1 receptor, also reduced trial, in which the dosages and formulation of metoprolol all-cause mortality and vascular events in the ACS trials tartrate have been heavily debated.59 The remaining three (RR 0.63, 95% CI 0.55 to 0.74, and RR 0.77, 95% CI 0.67 trials included only 472 patients and were not powered to 0.88, respectively). In addition, specific a1-receptor to detect a difference on mortality. The results of the blockers are not effective in patients with HF, also when COMET trial point to beneficial effects of carvedilol on combined with b1 blockers.63 reducing vascular events, which could also be explained Previous studies have suggested that the release of by antiadrenergic effects.60 norepinephrine is partly regulated by prejunctional

Consequently, we extended our analysis to placebo-controlled b2-adrenergic receptors. This implies that b1+2 blockers trials assessing the efficacy of the different b blockers, have a specific sympathoinhibitory effect that is less a method prone to potential biases. These studies have prominent in more selective b1 blockers. Indeed, b1+2 different designs and are heterogeneous, which impairs blockers reduce norepinephrine levels more effectively comparing these studies. Different types and dosages of compared with selective b1 blockers.14-16 Furthermore, b blockers were assessed, the study subjects differed, and increased (nor)epinephrine plasma levels enhance clinical outcomes were not clearly reported in every trial. coagulation activity by increased platelet activity and In addition, the b1 blocker trials were in general older than thrombin generation,13,64 and these prothrombotic effects the studies assessing the b1+2 blockers. However, b1+2 can be blocked by the b1+2 blocker propranolol but blockers studies showed far more consistency compared not by metoprolol or phentolamine, which points to a with b1 blockers, with a reduction of vascular events in all specific b2-adrenergic receptor mediated effect.13 Thus except one trial. Furthermore, the validity of our findings is b1+2 blockers not only reduce sympathetic activity more supported by the absence of heterogeneity among studies for effectively, but also the associated platelet activation and the major outcomes, and by the sensitivity analysis, where increments in coagulation factors.13,64-66 The net result the better efficacy ofb 1+2 blockers remained in high-quality could be a reduced prothrombotic state, thereby reducing studies. In addition, removing individual trials or analysing both arterial as venous thrombotic events. different types of b blockers did not affect our main results. In conclusion, this systematic review suggests that Five out of six studies that investigated the efficacy of b1+2 suppression of the b2-adrenergic receptor in addition blockers in patients with ACS, including the three largest to the b1 receptor may be more effective in reducing studies involving almost 8000 patients, found a reduction vascular events in patients with ACS and HF. The current in vascular events.28,35,57 literature is too heterogeneous to draw firm conclusions. Our aim was to assess the influence ofb 2-receptor blockade Nevertheless, the presumed antithrombotic effect of in addition to b1-receptor blockade. We therefore compared b2-adrenergic receptor blockers may call for additional b1 versus b1+2 blockers. Due to this classification we also studies assessing the beneficial effect of b1+2 blockers in included third-generation b blockers (such as carvedilol patients with ACS or HF.

De Peuter, et al. Beta blockers and vascular events.

october 2009, vol. 67, no 9 292 Acknowledgement 16. Kohno T, Yoshikawa T, Yoshizawa A, et al. Carvedilol exerts more potent antiadrenergic effect than metoprolol in heart failure. Cardiovasc Drugs Ther. 2005;19:347-55.

This study was financially supported by an unrestricted 17. Maclure M, Willett WC. Misinterpretation and misuse of the kappa grant of the Netherlands Heart Foundation (2006B180). statistic. Am J Epidemiol. 1987;126:161-9. 18. Hjalmarson A, Elmfeldt D, Herlitz J, et al. Effect on mortality of metoprolol in acute myocardial infarction. A double-blind randomised trial. Lancet. 1981;2:823-7. References 19. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials. 1996;17:1-12. 1. Anderson JL, Adams CD, Antman EM, et al. ACC/AHA 2007 Guidelines for the Management of Patients With Unstable Angina/Non ST-Elevation 20. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Myocardial Infarction. A Report of the American College of Cardiology/ Trials. 1986;7:177-88. American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 2002 Guidelines for the Management of 21. Sterne JA, Egger M, Smith GD. Systematic reviews in health care: Patients With Unstable Angina/Non ST-Elevation Myocardial Infarction). Investigating and dealing with publication and other biases in Circulation. 2007;116:e148-304. meta-analysis. BMJ. 2001;323:101-5.

2. Hunt SA. ACC/AHA 2005 guideline update for the diagnosis and 22. Manger Cats V, Capelle FJL, Lie KI, Durrer D. Effect of treatment with management of chronic heart failure in the adult: a report of the American 2x100 mg metoprolol on mortality in a single-center study with low College of Cardiology/American Heart Association Task Force on Practice placebo mortality-rate after infarction. Circulation. 1983;68(suppl.3):181. Guidelines (Writing Committee to Update the 2001 Guidelines for 23. Salathia KS, Barber JM, McIlmoyle EL, et al. Very early intervention the Evaluation and Management of Heart Failure). J Am Coll Cardiol. with metoprolol in suspected acute myocardial infarction. Eur Heart J. 2005;46:e1-82. 1985;6:190-8. 3. Swedberg K, Cleland J, Dargie H, et al. The Task Force for the Diagnosis 24. Anderson JL, Lutz JR, Gilbert EM, et al. A randomized trial of low-dose and Treatment of Chronic Heart Failure of the European Society of beta-blockade therapy for idiopathic dilated cardiomyopathy. Am J Cardiol. Cardiology. Guidelines for the diagnosis and treatment of chronic heart 1985;55:471-5. failure: executive summary (update 2005). Eur Heart J. 2005;26:1115-40. 25. Aronow WS, Ahn C, Kronzon I. Effect of propranolol versus no propranolol 4. Werf vdF, Ardissino D, Betriu A, et al. The Task Force on the Management on total mortality plus nonfatal myocardial infarction in older patients with of Acute Myocardial Infarction of the European Society of Cardiology. prior myocardial infarction, congestive heart failure, and left ventricular Management of acute myocardial infarction in patients presenting with ejection fraction > or = 40% treated with diuretics plus angiotensin- ST-segment elevation. Eur Heart J. 2003;24:28-66. converting enzyme inhibitors. Am J Cardiol. 1997;80:207-9. 5. Cohn JN, Levine TB, Olivari MT, et al. Plasma norepinephrine as a guide 26. Australia/New Zealand Heart Failure Research Collaborative Group. to prognosis in patients with chronic congestive heart failure. N Engl J Randomised, placebo-controlled trial of carvedilol in patients with congestive Med. 1984;311:819-23. heart failure due to ischaemic heart disease. Lancet. 1997;349:375-80. 6. Lopez-Sendon J, Swedberg K, McMurray J, et al. The Task Force 27. Basu S, Senior R, Raval U, et al. Beneficial effects of intravenous and oral on Beta-Blockers of the European Society of Cardiology. Expert carvedilol treatment in acute myocardial infarction. A placebo-controlled, consensus document on beta-adrenergic receptor blockers. Eur Heart randomized trial. Circulation. 1997;96:183-91. J. 2004;25:1341-62. 28. Beta-Blocker Heart Attack Trial Research Group. A randomized trial of 7. Freemantle N, Cleland J, Young P, Mason J, Harrison J. Beta Blockade propranolol in patients with acute myocardial infarction. I. Mortality after myocardial infarction: systematic review and meta regression results. JAMA. 1982;247:1707-14. analysis. BMJ. 1999;318:1730-7. 29. Bristow MR, Gilbert EM, Abraham WT, et al. Carvedilol produces 8. Lechat P, Packer M, Chalon S, et al. Clinical effects of beta-adrenergic dose-related improvements in left ventricular function and survival in blockade in chronic heart failure: a meta-analysis of double-blind, subjects with chronic heart failure. Circulation. 1996;94:2807-16. placebo-controlled, randomized trials. Circulation. 1998;98:1184-91. 30. CIBIS-II Investigators and Committees. The Cardiac Insufficiency 9. Packer M, Antonopoulos GV, Berlin JA, et al. Comparative effects of Bisoprolol Study II (CIBIS-II): a randomised trial. Lancet. 1999;353:9-13. carvedilol and metoprolol on left ventricular ejection fraction in heart failure: results of a meta-analysis. Am Heart J. 2001;141:899-907. 31. CIBIS Investigators and Committees. A randomized trial of beta-blockade in heart failure. The Cardiac Insufficiency Bisoprolol Study (CIBIS). 10. Poole-Wilson PA, Swedberg K, Cleland JG, et al. COMET Investigators. Circulation. 1994;90:1765-73. Comparison of carvedilol and metoprolol on clinical outcomes in patients with chronic heart failure in the Carvedilol Or Metoprolol European Trial 32. Cleland JG, Pennell DJ, Ray SG, et al. Carvedilol hibernating reversible (COMET): randomised controlled trial. Lancet. 2003;362:7-13. ischaemia trial: marker of success investigators. Myocardial viability as a determinant of the ejection fraction response to carvedilol in patients 11. Torp-Pedersen C, Poole-Wilson PA, Swedberg K, et al. COMET with heart failure (CHRISTMAS trial): randomised controlled trial. Lancet. Investigators. Effects of metoprolol and carvedilol on cause-specific 2003;362:14-21. mortality and morbidity in patients with chronic heart failure. Am Heart J. 2005;149:370-6. 33. Cohn JN, Fowler MB, Bristow MR, et al. The US Carvedilol Heart Failure Study Group. Safety and efficacy of carvedilol in severe heart failure. 12. Thrall G, Lane D, Carroll D, Lip GY. A systematic review of the effects J Card Fail. 1997;3:173-9. of acute psychological stress and physical activity on haemorheology, coagulation, fibrinolysis and platelet reactivity: Implications 34. Colucci WS, Packer M, Bristow MR, et al. The US Carvedilol Heart Failure for the pathogenesis of acute coronary syndromes. Thromb Res. Study Group. Carvedilol inhibits clinical progression in patients with mild 2007;120:819-47. symptoms of heart failure. Circulation. 1996;94:2800-6.

13. Von Kanel R, Dimsdale JE. Effects of sympathetic activation by adrenergic 35. Dargie HJ, The CAPRICORN Investigators. Effect of carvedilol on outcome infusions on hemostasis in vivo. Eur J Haematol. 2000;65:357-69. after myocardial infarction in patients with left-ventricular dysfunction: the CAPRICORN randomised trial. Lancet. 2001;357:1385-90. 14. Azevedo ER, Kubo T, Mak S, et al. Nonselective versus selective beta-adrenergic receptor blockade in congestive heart failure: differential 36. Edes I, Gasior Z, Wita K. Effects of nebivolol on left ventricular function effects on sympathetic activity. Circulation. 2001;104:2194-9. in elderly patients with chronic heart failure: results of the ENECA study. Eur J Heart Fail. 2005;7:631-9. 15. Herman RB, Jesudason PJ, Mustafa AM, et al. Differential effects of carvedilol and atenolol on plasma noradrenaline during exercise in 37. Figulla HR, Krzeminska-Pakula M, Wrabec K, et al. Betaxolol is equivalent to humans. Br J Clin Pharmacol. 2003;55:134-8. carvedilol in patients with heart failure NYHA II or III: result of a randomized multicenter trial (BETACAR Trial). Int J Cardiol. 2006;113:153-60.

De Peuter, et al. Beta blockers and vascular events.

october 2009, vol. 67, no 9 293 38. Flather MD, Shibata MC, Coats AJ, et al. The SENIORS investigators. 52. Waagstein F, Bristow MR, Swedberg K, et al. Metoprolol in Dilated Randomized trial to determine the effect of nebivolol on mortality and Cardiomyopathy (MDC) Trial Study Group. Beneficial effects of metoprolol cardiovascular hospital admission in elderly patients with heart failure in idiopathic dilated cardiomyopathy. Lancet. 1993;342:1441-6. (SENIORS). Eur Heart J. 2005;26:215-25. 53. Beta-Blocker Heart Attack Trial Research Group. A randomized trial of 39. Hansteen V, Moinichen E, Lorentsen E, et al. One year’s treatment with propranolol in patients with acute myocardial infarction. II. Morbidity propranolol after myocardial infarction: preliminary report of Norwegian results. JAMA. 1983;250:2814-9. multicentre trial. Br Med J (Clin Res Ed). 1982;284:155-60. 54. Gundersen T, Abrahamsen AM, Kjekshus J, Ronnevik PK, The Norwegian 40. Jonsson G, Abdelnoor M, Muller C, et al. A comparison of the two Multicentre Study Group. Timolol-related reduction in mortality and beta-blockers carvedilol and atenolol on left ventricular ejection fraction reinfarction in patients ages 65-75 years surviving acute myocardial and clinical endpoints after myocardial infarction. a single-centre, infarction. Circulation. 1982;66:1179-84. randomized study of 232 patients. Cardiology. 2005;103:148-55. 55. Hansteen V. Beta blockade after myocardial infarction: the Norwegian 41. Kukin ML, Kalman J, Charney RH, et al. Prospective, randomized propranolol study in high-risk patients. Circulation. 1983;67:I57-60. comparison of effect of long-term treatment with metoprolol or carvedilol on symptoms, exercise, ejection fraction, and oxidative stress in heart 56. Packer M, Fowler MB, Roecker EB, et al. The Carvedilol Prospective failure. Circulation. 1999;99:2645-51. Randomized Cumulative Survival (COPERNICUS) Study Group. Effect of carvedilol on the morbidity of patients with severe chronic heart failure: 42. Lopressor Intervention Trial Research Group. The Lopressor Intervention results of the carvedilol prospective randomized cumulative survival Trial: multicentre study of metoprolol in survivors of acute myocardial (COPERNICUS) study. Circulation. 2002;106:2194-9. infarction. Eur Heart J. 1987;8:1056-64. 57. Pedersen TR, The Norwegian Multicenter Study Group. The Norwegian 43. MERIT-HF Study Group. Effect of metoprolol CR/XL in chronic heart Multicenter Study of Timolol after Myocardial Infarction. Circulation. failure: Metoprolol CR/XL Randomised Intervention Trial in Congestive 1983;67:I49-I53. Heart Failure (MERIT-HF). Lancet. 1999;353:2001-7. 58. Bonet S, Agusti A, Arnau JM, et al. Beta-adrenergic blocking agents 44. Metra M, Giubbini R, Nodari S, et al. Differential effects of beta-blockers in heart failure: benefits of vasodilating and non-vasodilating agents in patients with heart failure: A prospective, randomized, double-blind according to patients’ characteristics: a meta-analysis of clinical trials. comparison of the long-term effects of metoprolol versus carvedilol. Arch Intern Med. 2000;160:621-7. Circulation. 2000;102:546-51. 59. Doggrell SA. Has COMET solved the controversy as to whether carvedilol 45. Olsson G, Rehnqvist N, Sjogren A, Erhardt L, Lundman T. Long-term is better than metoprolol in heart failure? Expert Opin Pharmacother. treatment with metoprolol after myocardial infarction: effect on 3 year 2004;5:205-8. mortality and morbidity. J Am Coll Cardiol. 1985;5:1428-37. 60. Packer M. Do beta-blockers prolong survival in heart failure only by 46. Packer M, Colucci WS, Sackner-Bernstein JD, et al. The PRECISE Study inhibiting the beta1-receptor? A perspective on the results of the COMET Group. Double-blind, placebo-controlled study of the effects of carvedilol trial. J Card Fail. 2003;9:429-43. in patients with moderate to severe heart failure: The PRECISE Trial. Circulation. 1996;94:2793-9. 61. de Boer RA, Voors AA, Van Veldhuisen DJ. Nebivolol: third-generation beta-blockade. Expert Opin Pharmacother. 2007;8:1539-50. 47. Packer M, Coats AJ, Fowler MB, et al. Effect of carvedilol on survival in severe chronic heart failure. N Engl J Med. 2001;344:1651-8. 62. Keating GM, Jarvis B. Carvedilol: a review of its use in chronic heart failure. Drugs. 2003;63:1697-741. 48. Palazzuoli A, Quatrini I, Vecchiato L, et al. Left ventricular diastolic function improvement by carvedilol therapy in advanced heart failure. 63. Massie BM. A comment on COMET: how to interpret a positive trial? J Cardiovasc Pharmacol. 2005;45:563-8. J Card Fail. 2003;9:425-8.

49. The Beta-Blocker Evaluation of Survival Trial Investigators. A trial of the 64. Grant PJ. Hormonal regulation of the acute haemostatic response to beta-blocker bucindolol in patients with advanced chronic heart failure. stress. Blood Coagul Fibrinolysis. 1990;1:299-306. N Engl J Med. 2001;344:1659-67. 65. Grignani G, Pacchiarini L, Zucchella M, et al. Effect of mental stress on 50. The Norwegian Multicenter Study Group. Timolol-induced reduction platelet function in normal subjects and in patients with coronary artery in mortality and reinfarction in patients surviving acute myocardial disease. Haemostasis. 1992;22:138-46. infarction. N Engl J Med 1981;304:801-7. 66. Hoppener MR, Kraaijenhagen RA, Hutten BA, et al. Beta-receptor 51. Torp-Pedersen C, Kober L, Ball S, et al. The incomplete bucindolol blockade decreases elevated plasma levels of factor VIII:C in patients with evaluation in acute myocardial infarction Trial (BEAT). Eur J Heart Fail. deep vein thrombosis. J Thromb Haemost. 2004;2:1316-20. 2002;4:495-9.

De Peuter, et al. Beta blockers and vascular events.

october 2009, vol. 67, no 9 294